This repository contains a Python script designed to navigate an Anki Vector robot through a predefined path, capturing images at various points and generating descriptive captions based on the content of the images. The script integrates a pre-trained Vision-Text model for image captioning and utilizes the Anki Vector SDK for robot control, providing a seamless and interactive user experience.
- Anki Vector Robot with SDK installed
- Python 3.6 or higher
- PyTorch
- NumPy
- PIL (Python Imaging Library)
- Hugging Face Transformers Library
-
Clone the repository:
git clone https://github.com/yourusername/anki-vector-navigation-captioning.git cd anki-vector-navigation-captioning -
Install the required packages:
pip install -r requirements.txt
- Ensure that the Anki Vector Robot is properly configured and connected to your development environment.
- The robot’s SDK should be installed and configured as per the official Anki Vector SDK documentation.
- Update the
data.jsonfile with the predefined poses for navigation. Each pose should contain X, Y coordinates and a direction angle.
-
Run the script:
python main.py
-
The script will initialize the Anki Vector Robot, capturing images and generating captions at each point in the predefined path.
-
Captions will be vocalized using the robot’s built-in text-to-speech functionality.
-
Movement details, including distances and directions, will be logged to the console for real-time monitoring.
- Robot Navigation: Navigate through a predefined path based on the poses specified in
data.json. - Image Captioning: Capture images using the robot’s camera and generate descriptive captions using a pre-trained Vision-Text model.
- Voice Feedback: Vocalize generated image captions for enhanced user interaction.
- Error Handling: Robust handling of
VectorTimeoutExceptionto ensure uninterrupted operation. - Real-time Logging: Console outputs for movement details and operational status.
This project is licensed under the MIT License - see the LICENSE.md file for details.